Experience from Asian centers in a named-patient-use program for afatinib in patients with advanced non-small-cell lung cancer who had progressed following prior therapies, including patients with uncommon EGFR mutations

Background This study evaluated outcomes among patients with advanced/metastatic non-small-cell lung cancer (NSCLC) treated at Asian centers participating in the global named-patient-use (NPU) program for afatinib. Methods Patients had progressed after initial benefit with erlotinib or gefitinib, and/or had an EGFR or HER2 mutation, had no other treatment options, and were ineligible for afatinib trials. The recommended starting dose of afatinib was 50 mg/day. Dose modifications were allowed, and afatinib was continued as long as deemed beneficial. Response and survival information was provided voluntarily. Safety reporting was mandatory. Results 2242 patients (26% aged ≥ 70 years, 96% with adenocarcinoma) received afatinib at centers in 10 Asian countries. Most were heavily pre-treated, including prior treatment with erlotinib or gefitinib. Of 1281 patients tested, 1240 had EGFR mutations (common: 1034/1101; uncommon: 117/1101). There were no new safety signals, the most common adverse events being rash and diarrhea. Objective response rate (ORR) was 24% overall (n = 431 with data available), 27% for patients with common EGFR mutations (n = 230) and 28% for those with uncommon mutations (n = 32); median time to treatment failure (TTF) in these groups was 7.6 months (n = 1550), 6.4 months (n = 692) and 8.4 months (n = 83), respectively. In patients with EGFR exon 20 insertions (n = 23) and HER2 mutations (n = 12), median TTF exceeded 12 months. Conclusions Patient outcomes in this study were similar to those reported in the analysis of the global NPU. Afatinib achieved clinical benefits in patients with refractory NSCLC. ORR and TTF were similar between patients with tumors harboring uncommon and common EGFR mutations.


Introduction
Afatinib is an irreversible inhibitor of the ErbB receptor family (EGFR [epidermal growth factor receptor]/ErbB1; HER2 [human epidermal growth factor receptor 2]/ErbB2; and HER4/ErbB4). In contrast to the first-generation EGFR tyrosine kinase inhibitors (TKIs) erlotinib and gefitinib, which bind reversibly to the ErbB1 receptor, afatinib covalently binds to all ErbB family receptors, blocking signaling and causing sustained inhibition of mitogenic activity [1,2]. Afatinib is approved in the European Union, USA, Canada, Switzerland, Australia, and several Asian, Latin American, and Middle Eastern countries as an oral, oncedaily tablet for patients with non-small-cell lung cancer (NSCLC) and activating EGFR mutations. In addition to the common mutations, exon 19 deletions (del19) and L858R substitutions, there is evidence that afatinib is active against some uncommon EGFR mutations, including L861Q, G719X, and S768I [3].
A global named-patient-use (NPU) program for afatinib was initiated in Germany and Australia in May 2010, for patients with advanced or metastatic NSCLC who had progressed after clinical benefit during previous treatment with erlotinib or gefitinib and/or had an activating EGFR/HER2 mutation, had exhausted all other treatments and were ineligible for an afatinib trial. The main objective of the program was to provide compassionate access to treatment for patients with no other established therapeutic options. The program continued until January 2016, by which time a total of 5636 patients had been treated in 49 countries on six continents. In an analysis of treatment outcomes in 3966 patients from 41 countries (excluding Taiwan), the median time to treatment failure (TTF) was 4.4 months and the objective response rate (ORR) was 23% [4]. Outcomes of patients treated in centers in specific countries [5][6][7][8], and with HER2 mutations [9] have also been described.
Here we present an analysis of treatment outcomes in patients who were treated at centers in 10 Asian countries. The large size of the NPU program made it possible to evaluate treatment outcomes in patients with both common and uncommon EGFR mutations. Understanding the influence of EGFR mutations is particularly important for patients in Asian countries, as EGFR mutations are prevalent among patients from this region [10].

Materials and methods
The design of the NPU has been reported previously [4]. Key details are summarized below.

Patients
Patients were eligible if they had advanced/metastatic NSCLC, had progressed after initially achieving clinical benefit (complete response [CR], partial response [PR], or stable disease [SD] lasting at least 6 months) during treatment with erlotinib or gefitinib and/or had an activating EGFR mutation or a HER2 mutation, had exhausted other treatment options and were ineligible to participate in an afatinib trial. Previous TKI therapy was not mandatory for all patients with confirmed mutations. Chemotherapy-naïve patients were eligible for inclusion if they were unfit to receive chemotherapy and were deemed ineligible to participate in an actively recruiting afatinib trial. The NPU program procedures (including enrollment criteria and treatment details) were adapted locally and approved in each region according to local regulations. The current analysis was conducted using data collected for patients treated at centers in Asian countries only.

Afatinib dose
The recommended starting dose of afatinib was 50 mg/ day, as used in the phase III LUX-Lung 1 study of afatinib following failure of prior erlotinib/gefitinib [11]. Lower starting doses (40 or 30 mg/day) were allowed at the discretion of the treating physician. Tolerability-guided dose modifications were also allowed, using 10-mg steps to a maximum of 50 mg/day and a minimum of 30 mg/day. Afatinib was continued as long as deemed beneficial by the treating physician.
Enrollment into the NPU program was terminated within each country once afatinib became commercially available locally; enrollment had ceased worldwide by January 2016. In some countries, patients were switched to commercially available afatinib provided by Boehringer Ingelheim; in others, patients continued to receive afatinib via the NPU program. In both cases, afatinib was continued as long as treatment was deemed beneficial by the treating physician.

Outcome measures
TTF was defined as the time from the date of initiation of afatinib to the date of discontinuation, switch to another drug, death, or the last available data, whichever occurred first. The ORR was defined as the proportion of patients with a recorded response outcome (CR, PR, SD, progressive disease [PD], or mixed response) who achieved a CR or PR. The disease control rate (DCR) was defined as a proportion of patients with response information available who achieved a CR, PR, or SD. There was no independent radiologic verification of responses, SD or PD.
Physicians were required to report safety-related information including all adverse events (AEs) leading to discontinuation of afatinib or deemed to be related to afatinib by the treating physician, and all serious AEs.

Data capture and analysis
Data collected during the program were provided voluntarily by the participating physicians and only safety reporting was mandatory.
No site monitoring, site audits, data cleaning, or structured data collection was conducted, except for the use of standard serious AE forms and reporting. Data were analyzed using SAS® software (Version 9.4, SAS Institute, Inc., NC, USA). The cut-off date for data analysis was January 18, 2016.
Baseline demographics and outcomes during treatment with afatinib were analyzed. Subgroup analyses were also conducted, including analyses of groups defined by the presence or absence of common and uncommon EGFR mutations, specific EGFR mutations, and HER2 mutations.

Ethics
All procedures were approved by the responsible ethics committees and the required country-specific and regional regulatory authorities were informed about the program.
The datasets generated and analyzed during the NPU program are available from the corresponding author on reasonable request.

Patient characteristics and treatment
As of the cut-off date, 2242 patients with NSCLC had received afatinib at treatment centers across 10 Asian countries (Table 1); 26% were aged ≥ 70 years. Of the 1924 patients with known tumor histology, 96% had adenocarcinoma. Most patients were heavily pre-treated; of the 2242 patients treated with afatinib, 2223 had received at least one line of previous therapy and 2202 had been treated with erlotinib and/or gefitinib. Sixty-two percent had previously received at least two lines of chemotherapy and 65% had received at least three lines of systemic therapy ( Table 2). Nineteen patients who were ineligible for chemotherapy received first-line afatinib at the request of the treating physicians on the basis that they considered it of potential benefit.
Twelve patients were HER2 mutation-positive, with no concurrent EGFR mutations. Seven had a specified HER2 mutation, all being p.A775_G776insYVMA insertions at nucleotide 2325 (Table 3).

Response to therapy, overall and according to mutation status
Information on response to afatinib was provided for 431 patients (19%), 78% of whom (335/431) had PR or SD; the ORR was 24% and the DCR was 78% (Table 4). Among patients with available information on both response and mutation status, ORR was 28% in patients with EGFR mutations (n = 267), 27% in those with common EGFR mutations (n = 230), and 28% in those with uncommon EGFR mutations (n = 32). In patients with G719X, L861Q, or S768I (n = 7), the ORR was 43% (Table 4). For patients with a specified HER2 mutation, 4/7 had response information available; all four had disease control (PR or SD) and one of the four had a PR (Table 4). In patients with a < 6-month interval between discontinuing previous EGFR TKI treatment and initiating afatinib (n = 586), the ORR was 23.5% and the DCR was 78.2%. In those patients with a < 12 month interval (n = 986), the ORR and DCR were 25.5 and 78.2%, respectively.

TTF, overall and according to mutation status and previous use of TKIs
Among the 1550 patients (69%) for whom TTF data were available, median TTF was 7.6 months (Table 5). Among those for whom information on both TTF and mutation status was available, median TTF was 7.2 months for patients with EGFR mutations (n = 834), 6.4 months for those with common EGFR mutations (n = 692) and 8.4 months for those with uncommon EGFR mutations (n = 83). In patients with G719X, L861Q, or S768I mutations (n = 28), median TTF was 7.8 months, but in those with EGFR ex20ins (n = 23) and HER2 mutations (n = 12), median TTF exceeded 12 months. Median TTF for patients who had previously received erlotinib or gefitinib was 8.7 months, and 8.2 months for those who had previously responded (CR/PR) to erlotinib (Table 5). In patients with < 6-/ < 12-month intervals between the discontinuation of prior EGFR TKI treatment and initiating afatinib, median TTF was 8.4 and 7.6 months, respectively.

Adverse events
The most frequently reported AEs were rash and diarrhea (Table 6).

Discussion
This analysis of patients from Asian countries involved in the afatinib NPU program revealed clinically meaningful ORRs and TTF in this heavily pretreated and resistant/ refractory advanced NSCLC patient population. Ninetyseven percent of patients with information on EGFR mutation status were EGFR mutation-positive, whereas 93% of the global NPU population were EGFR mutation-positive (2407/2595 patients, excluding patients from Taiwan) [4].
The findings of this study were generally similar to those of an analysis of the global NPU population [4]. This is an important finding as, without evidence, it cannot be assumed that patients from Asia respond similarly to those from other regions. However, median TTF for EGFR mutation-positive  Total a HER2 mutation-positive 12 (0.5 b ) Also EGFR mutation-positive 0/12 (0 b ) HER2 mutation specified 7/12 (58 f ) p.A775_G776insYVMA HER2 7/7 (100 g ) patients was 7.2 months for the Asian centers, compared with 4.3 months for the global NPU program, while the ORRs were 28% and 25%, respectively. The disparity in median TTF may have been influenced by differences between regions in the rates of reporting of key end dates, such as end of treatment and date of death. The median TTF for patients in this analysis of data from Asian centers may also have been influenced by the particularly prolonged TTF of patients in Taiwan 14.2 months (n = 273) [12]. When the NPU program began, investigators/clinicians in Taiwan had already gained experience in the use of afatinib, through their involvement in phase I and II trials. As a result, they may have been able to optimize their use of afatinib, potentially through more successful management of side effects, thereby allowing patients to remain on treatment for longer and to achieve greater clinical benefit. In contrast, in many other countries, physicians' first experience with afatinib was within the NPU.
The clinical benefits of afatinib appeared to be similar between patients with NSCLC tumors harboring common or uncommon EGFR mutations. The median TTF in the subgroup with uncommon EGFR mutations was 8.4 months, versus 6.4 months in the subgroup with common EGFR mutations; ORR was 28% and 27%, respectively. In the global NPU program, TTF was 4.3 months in both subgroups [4], while ORR was 26% and 25%, respectively [4].
In the current analysis, among the 28 patients with G719X, L861Q, or S768I mutations, median TTF was 7.8 months, compared with 4.7 months for patients with these mutations in the global NPU population (n = 77; [4]). The ORR in the current analysis was 43% (3/7 patients), compared with 30% (8/27 patients) in the global NPU analysis [4]. The difference in median TTF between patients with G719X, L861Q, or S768I mutations in the Asian subgroup and the global NPU program may have been particularly influenced by the median of 21.1 months among patients from Taiwan [12], who were not included in the global NPU analysis [4]. However, the prolonged outcomes reported here for patients with G719X, L861Q, or S768I mutations are in agreement with data from the LUX-Lung 2, 3, and 6 studies; for patients with G719X, L861Q or S768I, ORR ranged from 56 to 100%, median PFS 8-15 months, and median OS was 17 months-not estimated. The authors concluded that afatinib has activity against NSCLC tumors harboring these types of uncommon EGFR mutations [3].
Compared with patients with other EGFR mutations, patients with EGFR ex20ins have been reported to respond relatively infrequently to first-generation TKIs [13][14][15]  and have poorer survival outcomes than those with common EGFR mutations [16]. One review of 84 patients with ex20ins (based on eight small series of between 2 and 25 patients) treated with gefitinib or erlotinib found a response rate of 11% [17]. In another study, median TTF for 8 patients with EGFR ex20ins taking erlotinib for advanced disease was 2.4 months, compared with 5.9 months for 17 patients taking combination chemotherapy [16]. In the LUX-Lung 2, 3, and 6 trials, responses to afatinib were least common in patients with ex20ins, with two of 23 patients (8.7%) having an objective response [3]. Afatinib did not confer clinical benefit in these patients. Median PFS was 2.7 months (n = 23), which was similar to the PFS of 2.9 months in 14 patients with de-novo T790M resistance mutations [3]. Based on the evidence to date, NSCLC patients harboring EGFR ex20ins have been described as a unique subset, for whom there are no effective approved targeted therapies [15]. In terms of response to treatment, our findings for patients with ex20ins are similar to previous reports. In this study, 1/5 patients with ex20ins (20%) exhibited PR with afatinib, with 7/20 patients (35%) in the analysis of the global NPU program [4]. The median TTF for 23 patients with EGFR ex20ins in this study, however, was 18.9 months, longer than for any other subgroup in this analysis, and much longer than the median of 3.6 months for patients with ex20ins in the global NPU program (n = 57) [4]. This disparity may be explained, in part, by the presence of clinical and biochemical heterogeneity between different types of EGFR ex20ins, which actually represent a broad subtype of EGFR mutations. Kosaka et al. [15] identified at least 19 different types of ex20ins [15], which differed markedly in their susceptibility to afatinib in vitro. While most ex20ins are resistant to EGFR inhibitors, one appears to confer sensitivity to erlotinib (A763_Y764insFQEA) [18,19]. There is also evidence that the nature of the ex20ins can influence sensitivity to afatinib. A patient in the NPU program with an EGFR ex20ins (A767_S768insSVA tandem duplication) was maintained at minimal PD for 54 months and received afatinib for 36 months [20]; this raises the possibility that some patients with EGFR ex20ins may obtain benefit from afatinib in this setting. Recently, Liu et al. [21] identified ex20ins in 2% of Chinese patients with lung cancer (171/7,520); detailed survival information during afatinib monotherapy was available for 19 patients. Those harboring G778_P780dup (G778) had longer PFS (median 10 vs. 3.3 months, p = 0.32) and OS (median 19.7 vs. 7 months, p = 0.16) than those with other ex20ins [21]. A limitation of the current analysis is that ex20ins type was not identified for most patients. This may be important for the correct interpretation of therapeutic outcomes in patients with ex20ins and should be considered in future studies. All 12 patients who were HER2 mutation-positive were treated at centers in Taiwan, and all specified HER2 mutations were in exon 20. Most HER2 mutations are in exon 20, occurring in 2-4% of patients with NSCLC [22][23][24][25].
In the current analysis, median TTF in this subgroup was 12.2 months, and 12.4 months in the 7 patients with the p.A775_G776insYVMA HER2 mutation. For the 28 HER2 mutation-positive patients in the global NPU, the median TTF was 2.9 months, but eight of the 28 patients (29%) had TTF > 1 year. The most frequent HER2 mutation was p.A775_G776insYVMA (insertion at nucleotide 2325), which was identified in 10 of the 12 patients with specified HER2 mutations (83%). The median TTF for these patients was 9.6 months [9].
During the NPU program, the third-generation EGFR TKI osimertinib was approved for patients with EGFR T790M-mutation-positive NSCLC, following acquired resistance to EGFR TKIs. In a phase II study, the ORR of osimertinib in pretreated T790M-positive NSCLC patients was 62% [26], higher than reported here (25%). Osimertinib has also been reported to achieve a high response rate in patients with the uncommon L861Q and G719A/C/D/S/X EGFR mutations [27]. The effects of osimertinib treatment in patients with other uncommon EGFR mutations have not been reported but, given its specificity for EGFR/ErbB1, unlike afatinib, osimertinib would not be expected to confer benefits in patients with HER2 mutations.
The safety data observed across the Asian centers were consistent with those reported from non-Asian centers in the NPU program [4,7,8] and with the known safety profile of afatinib. Several of the more frequently reported events were not considered to be side effects of afatinib but were symptoms or consequences of advanced-stage lung cancer or disease progression (e.g. fatigue/asthenia, pneumonia, pleural effusion, and cough).
The current analysis has some limitations. The data generated by the NPU program were derived from patients treated in accordance with local clinical practices, but data collection relied on voluntary reporting of patient information by the investigators. Consequently, the proportions of patients with data available on tumor histology, EGFR mutation type, response and TTF were relatively low. Also, physicians may have been less spontaneous in providing information on response if their patients did not show a significant therapeutic response; this could have led to under-reporting of patients who did worse, and overestimation of ORR and TTF. Centralized radiographic confirmation of tumor response/SD was not available. The data may also have been influenced by failures to report certain end dates, such as end of treatment and date of death. Additionally, AEs may have been under-reported in this real-world setting compared with the rates of reporting generally seen in clinical trials. Finally, information on EGFR mutation status was available for 57% of the patients, but the validation of the results by a central laboratory was not available. No information was available on the timing of the mutation analysis, particularly in relation to the timing of first-generation TKI therapy. Nevertheless, this afatinib NPU program provided practical data on the safety and efficacy of afatinib in the setting of local clinical practices in diverse Asian countries, while facilitating compassionate use of afatinib in a large group of NSCLC patients who had exhausted all other treatment options.

Conclusions
The findings of the current analysis of data from Asian centers and those of the global NPU [4] were generally similar. Afatinib achieved clinical benefits in patients with refractory NSCLC, both with common and uncommon EGFR mutations. In patients with tumors harboring uncommon EGFR mutations, estimates of ORR and TTF were similar, if not superior, to those in patients with common EGFR mutations. The variation in the efficacy of afatinib treatment in NSCLC harboring different uncommon mutations remains an important area for future research.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.